Patients with recent homelessness were designated as awaiting alternative care at more than four times the rate of matched comparators in Ontario

A population-based matched cohort study in Ontario, Canada, covered 51,377 hospital patients aged 16 and over who were discharged between April 2022 and March 2024 Primary Study (Richard et al., PMID 42467431). It compared 11,168 patients with a recent experience of homelessness, identified through a validated definition in health administrative data, with 40,209 patients from the neighbourhoods with the lowest level of material resources, matched four to one on age, sex, month of admission, health region and comorbidity. The outcome was designation as alternate level of care (ALC), the status given to a patient who is occupying a bed while no longer needing the intensity of care a hospital provides.

ALC designation was 9.1% (1,014 of 11,168) among patients with recent homelessness against 2.0% (810 of 40,209) among matched comparators, a rate ratio of 4.51 (95% confidence interval 4.11–4.95). Among patients who were designated, those with recent homelessness spent longer in that status, with an incidence rate ratio of 1.22 (1.09–1.37) and a difference of 4.7 days (2.1–7.4). The gap varied widely by health region, from a rate ratio of 2.39 to 7.88, which the authors attribute to differences in local capacity and supports.

Because the comparators were drawn from the most materially deprived neighbourhoods, the excess is beyond what neighbourhood deprivation alone would produce. The authors read it as structural barriers to safe discharge locations and gaps in transitional and post-acute care, and name medical respite and supportive housing as the interventions that would address it; the study itself did not test any intervention.

The limits are material. It is observational, based on administrative data, and Canadian: no UK comparator was identified in the source material, and the care arrangements that produce a delayed-discharge designation differ between systems. The patients were also younger than a typical frailty cohort (mean age about 49), so the result bears on the distribution of delay across the population, not on older people’s discharge specifically. Its value is the quantification: delay is not spread evenly, and an average length-of-delay figure would conceal the group that carries most of it.